Face Generation

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data

You'll be using two datasets in this project:

  • MNIST
  • CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

In [1]:
data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)
Downloading mnist: 9.92MB [00:00, 46.3MB/s]                            
Extracting mnist: 100%|██████████| 60.0K/60.0K [00:11<00:00, 5.25KFile/s]
Downloading celeba: 1.44GB [01:29, 16.2MB/s]                               
Extracting celeba...

Explore the Data

MNIST

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

In [2]:
show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')
Out[2]:
<matplotlib.image.AxesImage at 0x7f4e91b22ac8>

CelebA

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

In [3]:
show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))
Out[3]:
<matplotlib.image.AxesImage at 0x7f4e91a4f9b0>

Preprocess the Data

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network

You'll build the components necessary to build a GANs by implementing the following functions below:

  • model_inputs
  • discriminator
  • generator
  • model_loss
  • model_opt
  • train

Check the Version of TensorFlow and Access to GPU

This will check to make sure you have the correct version of TensorFlow and access to a GPU

In [4]:
"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))
TensorFlow Version: 1.1.0
Default GPU Device: /gpu:0

Input

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

  • Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
  • Z input placeholder with rank 2 using z_dim.
  • Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

In [5]:
import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    inputs_real = tf.placeholder(tf.float32, (None, image_width, image_height, image_channels), name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')
    learning_rate = tf.placeholder(tf.float32, name='learning_rate')

    return inputs_real, inputs_z, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)
Tests Passed

Discriminator

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

In [6]:
def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    alpha = 0.2

    with tf.variable_scope('discriminator', reuse=reuse):
        # Input layer is 28x28x3
        x1 = tf.layers.conv2d(images, 64, 5, strides=2, padding='same')
        relu1 = tf.maximum(alpha * x1, x1)
        # 14x14x64

        x2 = tf.layers.conv2d(relu1, 128, 5, strides=2, padding='same')
        bn2 = tf.layers.batch_normalization(x2, training=True)
        relu2 = tf.maximum(alpha * bn2, bn2)
        # 7x7x128

        x3 = tf.layers.conv2d(relu2, 256, 5, strides=2, padding='same')
        bn3 = tf.layers.batch_normalization(x3, training=True)
        relu3 = tf.maximum(alpha * bn3, bn3)
        # 4x4x256

        # Flatten it 4096
        flat = tf.reshape(relu3, (-1, 4 * 4 * 256))
        logits = tf.layers.dense(flat, 1)
        out = tf.sigmoid(logits)

        return out, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)
Tests Passed

Generator

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

In [7]:
def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    alpha = 0.2

    with tf.variable_scope('generator', reuse=not is_train):
        # First fully connected layer
        x1 = tf.layers.dense(z, 4 * 4 * 512)
        # Reshape it to start the convolutional stack
        x1 = tf.reshape(x1, (-1, 4, 4, 512))
        x1 = tf.layers.batch_normalization(x1, training=is_train)
        x1 = tf.maximum(alpha * x1, x1)
        # 4x4x512 now

        x2 = tf.layers.conv2d_transpose(x1, 256, 5, strides=2, padding='same')
        x2 = tf.layers.batch_normalization(x2, training=is_train)
        x2 = tf.maximum(alpha * x2, x2)
        # 8x8x256 now

        x3 = tf.layers.conv2d_transpose(x2, 128, 7, strides=1, padding='valid')
        x3 = tf.layers.batch_normalization(x3, training=is_train)
        x3 = tf.maximum(alpha * x3, x3)
        # 14x14x128 now

        # Output layer
        logits = tf.layers.conv2d_transpose(x3, out_channel_dim, 5, strides=2, padding='same')
        # 28x28x3 now

        out = tf.tanh(logits)

        return out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)
Tests Passed

Loss

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

  • discriminator(images, reuse=False)
  • generator(z, out_channel_dim, is_train=True)
In [8]:
def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    g_model = generator(input_z, out_channel_dim)
    d_model_real, d_logits_real = discriminator(input_real)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)

    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real)))
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    d_loss = d_loss_real + d_loss_fake

    return d_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)
Tests Passed

Optimization

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

In [9]:
def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)

    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)
Tests Passed

Neural Network Training

Show Output

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

In [10]:
"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train

Implement train to build and train the GANs. Use the following functions you implemented:

  • model_inputs(image_width, image_height, image_channels, z_dim)
  • model_loss(input_real, input_z, out_channel_dim)
  • model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

In [11]:
class GAN:
    def __init__(self, real_size, z_size, learning_rate, beta1=0.5):
        tf.reset_default_graph()

        self.out_channel_dim = real_size[2]

        self.input_real, self.input_z, _ = model_inputs(real_size[0], real_size[1], real_size[2], z_size)

        self.d_loss, self.g_loss = model_loss(self.input_real, self.input_z, self.out_channel_dim)

        self.d_opt, self.g_opt = model_opt(self.d_loss, self.g_loss, learning_rate, beta1)
        
        
def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    print_every = 10
    show_every = 20

    net = GAN(data_shape[1:], z_dim, learning_rate, beta1=beta1)

    saver = tf.train.Saver()

    samples, losses = [], []
    steps = 0

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for e in range(epoch_count):
            for batch_images in get_batches(batch_size):
                steps += 1

                # batch_images = batch_images.reshape((batch_size, 28, 28, 1))
                batch_images = batch_images * 2

                # Sample random noise for G
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_dim))

                # Run optimizers
                _ = sess.run(net.d_opt, feed_dict={net.input_real: batch_images, net.input_z: batch_z})
                _ = sess.run(net.g_opt, feed_dict={net.input_z: batch_z, net.input_real: batch_images})

                if steps % print_every == 0:
                    # At the end of each epoch, get the losses and print them out
                    train_loss_d = net.d_loss.eval({net.input_z: batch_z, net.input_real: batch_images})
                    train_loss_g = net.g_loss.eval({net.input_z: batch_z})

                    print("Epoch {}/{}...".format(e + 1, epoch_count),
                          "Discriminator Loss: {:.4f}...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))
                    # Save losses to view after training
                    losses.append((train_loss_d, train_loss_g))

                if steps % show_every == 0:
                    show_generator_output(sess, 9, net.input_z, net.out_channel_dim, data_image_mode)
                    # gen_samples = sess.run(
                    #     generator(net.input_z, net.out_channel_dim, is_train=False),
                    #     feed_dict={net.input_z: sample_z})
                    # samples.append(gen_samples)
                    # view_samples("{}".format(steps), -1, samples, 6, 12, figsize=figsize)
                    # plt.show()

        saver.save(sess, './checkpoints/generator.ckpt')

    with open('samples.pkl', 'wb') as f:
        pkl.dump(samples, f)

    return losses, samples
                
                

MNIST

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

In [12]:
batch_size = 128

z_dim = 100
learning_rate = 0.0002
beta1 = 0.5



"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)
Epoch 1/2... Discriminator Loss: 0.1662... Generator Loss: 2.2687
Epoch 1/2... Discriminator Loss: 0.0703... Generator Loss: 3.3830
Epoch 1/2... Discriminator Loss: 0.0646... Generator Loss: 3.0979
Epoch 1/2... Discriminator Loss: 0.0157... Generator Loss: 4.5777
Epoch 1/2... Discriminator Loss: 0.0419... Generator Loss: 5.7393
Epoch 1/2... Discriminator Loss: 0.7479... Generator Loss: 0.7405
Epoch 1/2... Discriminator Loss: 0.0924... Generator Loss: 3.6406
Epoch 1/2... Discriminator Loss: 0.1944... Generator Loss: 12.5496
Epoch 1/2... Discriminator Loss: 0.6705... Generator Loss: 0.8006
Epoch 1/2... Discriminator Loss: 0.1389... Generator Loss: 3.5668
Epoch 1/2... Discriminator Loss: 0.5828... Generator Loss: 1.2224
Epoch 1/2... Discriminator Loss: 0.1774... Generator Loss: 3.3530
Epoch 1/2... Discriminator Loss: 0.2256... Generator Loss: 2.5277
Epoch 1/2... Discriminator Loss: 0.4438... Generator Loss: 1.4086
Epoch 1/2... Discriminator Loss: 0.2764... Generator Loss: 2.5376
Epoch 1/2... Discriminator Loss: 0.2400... Generator Loss: 2.8132
Epoch 1/2... Discriminator Loss: 0.5176... Generator Loss: 3.9651
Epoch 1/2... Discriminator Loss: 0.3891... Generator Loss: 1.9903
Epoch 1/2... Discriminator Loss: 1.0885... Generator Loss: 3.3756
Epoch 1/2... Discriminator Loss: 0.4741... Generator Loss: 4.6683
Epoch 1/2... Discriminator Loss: 1.5649... Generator Loss: 0.3312
Epoch 1/2... Discriminator Loss: 0.9477... Generator Loss: 1.5991
Epoch 1/2... Discriminator Loss: 0.5617... Generator Loss: 2.2773
Epoch 1/2... Discriminator Loss: 0.6754... Generator Loss: 1.4384
Epoch 1/2... Discriminator Loss: 0.4539... Generator Loss: 2.6134
Epoch 1/2... Discriminator Loss: 0.6261... Generator Loss: 1.1352
Epoch 1/2... Discriminator Loss: 0.8544... Generator Loss: 1.0130
Epoch 1/2... Discriminator Loss: 0.9317... Generator Loss: 1.2149
Epoch 1/2... Discriminator Loss: 0.8877... Generator Loss: 0.9632
Epoch 1/2... Discriminator Loss: 0.6568... Generator Loss: 1.5917
Epoch 1/2... Discriminator Loss: 0.5944... Generator Loss: 1.9821
Epoch 1/2... Discriminator Loss: 0.5812... Generator Loss: 2.0757
Epoch 1/2... Discriminator Loss: 0.6714... Generator Loss: 1.3731
Epoch 1/2... Discriminator Loss: 1.5162... Generator Loss: 3.3934
Epoch 1/2... Discriminator Loss: 0.7776... Generator Loss: 1.1985
Epoch 1/2... Discriminator Loss: 0.8503... Generator Loss: 0.9660
Epoch 1/2... Discriminator Loss: 0.9031... Generator Loss: 2.2960
Epoch 1/2... Discriminator Loss: 0.8467... Generator Loss: 1.1296
Epoch 1/2... Discriminator Loss: 0.9596... Generator Loss: 1.8104
Epoch 1/2... Discriminator Loss: 0.8954... Generator Loss: 1.1874
Epoch 1/2... Discriminator Loss: 1.1100... Generator Loss: 0.8201
Epoch 1/2... Discriminator Loss: 0.9327... Generator Loss: 0.9040
Epoch 1/2... Discriminator Loss: 1.0669... Generator Loss: 0.7861
Epoch 1/2... Discriminator Loss: 0.8655... Generator Loss: 0.9145
Epoch 1/2... Discriminator Loss: 0.9538... Generator Loss: 1.7771
Epoch 1/2... Discriminator Loss: 0.8661... Generator Loss: 1.2909
Epoch 2/2... Discriminator Loss: 1.1132... Generator Loss: 0.6403
Epoch 2/2... Discriminator Loss: 0.8968... Generator Loss: 1.4708
Epoch 2/2... Discriminator Loss: 0.9920... Generator Loss: 0.7142
Epoch 2/2... Discriminator Loss: 1.1706... Generator Loss: 0.5495
Epoch 2/2... Discriminator Loss: 0.9465... Generator Loss: 2.1977
Epoch 2/2... Discriminator Loss: 0.7231... Generator Loss: 1.1477
Epoch 2/2... Discriminator Loss: 0.9958... Generator Loss: 0.7067
Epoch 2/2... Discriminator Loss: 0.7699... Generator Loss: 1.1124
Epoch 2/2... Discriminator Loss: 0.8552... Generator Loss: 1.8563
Epoch 2/2... Discriminator Loss: 0.8243... Generator Loss: 1.3579
Epoch 2/2... Discriminator Loss: 0.8990... Generator Loss: 2.1239
Epoch 2/2... Discriminator Loss: 0.8275... Generator Loss: 0.9670
Epoch 2/2... Discriminator Loss: 0.8803... Generator Loss: 0.9142
Epoch 2/2... Discriminator Loss: 0.8010... Generator Loss: 1.5983
Epoch 2/2... Discriminator Loss: 0.8091... Generator Loss: 0.9090
Epoch 2/2... Discriminator Loss: 0.8241... Generator Loss: 1.8536
Epoch 2/2... Discriminator Loss: 0.7780... Generator Loss: 1.4196
Epoch 2/2... Discriminator Loss: 0.8506... Generator Loss: 0.9881
Epoch 2/2... Discriminator Loss: 0.9481... Generator Loss: 0.7582
Epoch 2/2... Discriminator Loss: 0.9489... Generator Loss: 0.7699
Epoch 2/2... Discriminator Loss: 1.2158... Generator Loss: 0.4908
Epoch 2/2... Discriminator Loss: 0.8150... Generator Loss: 1.7229
Epoch 2/2... Discriminator Loss: 1.3451... Generator Loss: 0.3863
Epoch 2/2... Discriminator Loss: 0.8043... Generator Loss: 1.2838
Epoch 2/2... Discriminator Loss: 0.9979... Generator Loss: 2.1842
Epoch 2/2... Discriminator Loss: 0.8064... Generator Loss: 1.0451
Epoch 2/2... Discriminator Loss: 1.1828... Generator Loss: 0.5453
Epoch 2/2... Discriminator Loss: 0.9369... Generator Loss: 0.7288
Epoch 2/2... Discriminator Loss: 0.7948... Generator Loss: 1.0719
Epoch 2/2... Discriminator Loss: 0.9413... Generator Loss: 1.9681
Epoch 2/2... Discriminator Loss: 0.8027... Generator Loss: 1.1686
Epoch 2/2... Discriminator Loss: 0.9618... Generator Loss: 1.7054
Epoch 2/2... Discriminator Loss: 0.7724... Generator Loss: 1.4545
Epoch 2/2... Discriminator Loss: 0.8748... Generator Loss: 0.9100
Epoch 2/2... Discriminator Loss: 0.7982... Generator Loss: 1.1813
Epoch 2/2... Discriminator Loss: 0.8277... Generator Loss: 1.5171
Epoch 2/2... Discriminator Loss: 0.8397... Generator Loss: 1.7738
Epoch 2/2... Discriminator Loss: 0.8796... Generator Loss: 1.6170
Epoch 2/2... Discriminator Loss: 1.0078... Generator Loss: 0.7693
Epoch 2/2... Discriminator Loss: 1.2669... Generator Loss: 2.5984
Epoch 2/2... Discriminator Loss: 1.0516... Generator Loss: 0.6055
Epoch 2/2... Discriminator Loss: 0.7736... Generator Loss: 1.0035
Epoch 2/2... Discriminator Loss: 0.9753... Generator Loss: 0.6359
Epoch 2/2... Discriminator Loss: 0.7769... Generator Loss: 1.2869
Epoch 2/2... Discriminator Loss: 0.8050... Generator Loss: 1.5016
Epoch 2/2... Discriminator Loss: 0.7922... Generator Loss: 0.9028
Epoch 2/2... Discriminator Loss: 0.8909... Generator Loss: 0.7397
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/usr/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py in get_controller(self, default)
   3624       self.stack.append(default)
-> 3625       yield default
   3626     finally:

<ipython-input-12-8557329b327d> in <module>()
     16     train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
---> 17           mnist_dataset.shape, mnist_dataset.image_mode)

<ipython-input-11-779a49d31bd0> in train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode)
     71 
---> 72         saver.save(sess, './checkpoints/generator.ckpt')
     73 

/usr/lib/python3.4/dist-packages/tensorflow/python/training/saver.py in save(self, sess, save_path, global_step, latest_filename, meta_graph_suffix, write_meta_graph, write_state)
   1381       raise ValueError(
-> 1382           "Parent directory of {} doesn't exist, can't save.".format(save_path))
   1383 

ValueError: Parent directory of ./checkpoints/generator.ckpt doesn't exist, can't save.

During handling of the above exception, another exception occurred:

IndexError                                Traceback (most recent call last)
<ipython-input-12-8557329b327d> in <module>()
     15 with tf.Graph().as_default():
     16     train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
---> 17           mnist_dataset.shape, mnist_dataset.image_mode)

/usr/lib64/python3.4/contextlib.py in __exit__(self, type, value, traceback)
     75                 value = type()
     76             try:
---> 77                 self.gen.throw(type, value, traceback)
     78                 raise RuntimeError("generator didn't stop after throw()")
     79             except StopIteration as exc:

/usr/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py in get_controller(self, default)
   3626     finally:
   3627       if self._enforce_nesting:
-> 3628         if self.stack[-1] is not default:
   3629           raise AssertionError(
   3630               "Nesting violated for default stack of %s objects"

IndexError: list index out of range

CelebA

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

In [15]:
batch_size = 128

z_dim = 100
learning_rate = 0.0002
beta1 = 0.5


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)
Epoch 1/2... Discriminator Loss: 0.4784... Generator Loss: 1.7669
Epoch 1/2... Discriminator Loss: 0.1973... Generator Loss: 2.9355
Epoch 1/2... Discriminator Loss: 0.1336... Generator Loss: 2.8210
Epoch 1/2... Discriminator Loss: 0.2661... Generator Loss: 8.5356
Epoch 1/2... Discriminator Loss: 3.1363... Generator Loss: 0.0461
Epoch 1/2... Discriminator Loss: 0.6028... Generator Loss: 8.6458
Epoch 1/2... Discriminator Loss: 0.7167... Generator Loss: 0.8883
Epoch 1/2... Discriminator Loss: 0.3236... Generator Loss: 7.6249
Epoch 1/2... Discriminator Loss: 0.5997... Generator Loss: 1.4347
Epoch 1/2... Discriminator Loss: 0.2766... Generator Loss: 3.0884
Epoch 1/2... Discriminator Loss: 0.2318... Generator Loss: 2.5384
Epoch 1/2... Discriminator Loss: 0.4885... Generator Loss: 2.1580
Epoch 1/2... Discriminator Loss: 0.2962... Generator Loss: 2.5585
Epoch 1/2... Discriminator Loss: 0.2703... Generator Loss: 3.4827
Epoch 1/2... Discriminator Loss: 0.2367... Generator Loss: 2.3179
Epoch 1/2... Discriminator Loss: 0.1987... Generator Loss: 2.4101
Epoch 1/2... Discriminator Loss: 0.2200... Generator Loss: 2.5405
Epoch 1/2... Discriminator Loss: 0.1619... Generator Loss: 2.5572
Epoch 1/2... Discriminator Loss: 0.2414... Generator Loss: 2.3468
Epoch 1/2... Discriminator Loss: 0.1375... Generator Loss: 2.9947
Epoch 1/2... Discriminator Loss: 0.1874... Generator Loss: 2.3938
Epoch 1/2... Discriminator Loss: 0.4007... Generator Loss: 1.7500
Epoch 1/2... Discriminator Loss: 0.6620... Generator Loss: 1.1450
Epoch 1/2... Discriminator Loss: 0.6504... Generator Loss: 1.0913
Epoch 1/2... Discriminator Loss: 0.3473... Generator Loss: 2.1339
Epoch 1/2... Discriminator Loss: 1.3929... Generator Loss: 0.4373
Epoch 1/2... Discriminator Loss: 0.5289... Generator Loss: 2.3957
Epoch 1/2... Discriminator Loss: 0.4447... Generator Loss: 2.0119
Epoch 1/2... Discriminator Loss: 0.4971... Generator Loss: 2.1223
Epoch 1/2... Discriminator Loss: 1.0270... Generator Loss: 1.3159
Epoch 1/2... Discriminator Loss: 0.7755... Generator Loss: 1.3250
Epoch 1/2... Discriminator Loss: 1.0562... Generator Loss: 0.8913
Epoch 1/2... Discriminator Loss: 1.0663... Generator Loss: 0.9890
Epoch 1/2... Discriminator Loss: 1.1694... Generator Loss: 0.7072
Epoch 1/2... Discriminator Loss: 1.2861... Generator Loss: 0.7699
Epoch 1/2... Discriminator Loss: 1.1039... Generator Loss: 1.2055
Epoch 1/2... Discriminator Loss: 1.3150... Generator Loss: 1.1773
Epoch 1/2... Discriminator Loss: 1.0914... Generator Loss: 1.2166
Epoch 1/2... Discriminator Loss: 1.6478... Generator Loss: 0.7480
Epoch 1/2... Discriminator Loss: 0.9452... Generator Loss: 1.8891
Epoch 1/2... Discriminator Loss: 0.9525... Generator Loss: 1.1491
Epoch 1/2... Discriminator Loss: 1.1497... Generator Loss: 1.2267
Epoch 1/2... Discriminator Loss: 1.6184... Generator Loss: 0.6550
Epoch 1/2... Discriminator Loss: 1.1676... Generator Loss: 0.8493
Epoch 1/2... Discriminator Loss: 1.1767... Generator Loss: 0.9331
Epoch 1/2... Discriminator Loss: 1.2045... Generator Loss: 0.8560
Epoch 1/2... Discriminator Loss: 1.2785... Generator Loss: 0.8805
Epoch 1/2... Discriminator Loss: 1.5512... Generator Loss: 0.9598
Epoch 1/2... Discriminator Loss: 1.1051... Generator Loss: 0.9198
Epoch 1/2... Discriminator Loss: 0.8990... Generator Loss: 1.0324
Epoch 1/2... Discriminator Loss: 0.9315... Generator Loss: 0.9764
Epoch 1/2... Discriminator Loss: 1.0049... Generator Loss: 0.9500
Epoch 1/2... Discriminator Loss: 0.9024... Generator Loss: 1.0240
Epoch 1/2... Discriminator Loss: 1.3067... Generator Loss: 0.8696
Epoch 1/2... Discriminator Loss: 1.1821... Generator Loss: 0.7229
Epoch 1/2... Discriminator Loss: 1.0521... Generator Loss: 0.9303
Epoch 1/2... Discriminator Loss: 0.8813... Generator Loss: 1.0613
Epoch 1/2... Discriminator Loss: 0.9738... Generator Loss: 1.1545
Epoch 1/2... Discriminator Loss: 1.2655... Generator Loss: 0.8849
Epoch 1/2... Discriminator Loss: 1.3356... Generator Loss: 1.0985
Epoch 1/2... Discriminator Loss: 1.0263... Generator Loss: 1.0053
Epoch 1/2... Discriminator Loss: 0.8757... Generator Loss: 1.6342
Epoch 1/2... Discriminator Loss: 0.8504... Generator Loss: 1.7858
Epoch 1/2... Discriminator Loss: 1.1426... Generator Loss: 0.7066
Epoch 1/2... Discriminator Loss: 0.9916... Generator Loss: 0.9309
Epoch 1/2... Discriminator Loss: 0.8322... Generator Loss: 1.2996
Epoch 1/2... Discriminator Loss: 0.8957... Generator Loss: 1.1253
Epoch 1/2... Discriminator Loss: 1.3967... Generator Loss: 1.0354
Epoch 1/2... Discriminator Loss: 1.0384... Generator Loss: 1.2009
Epoch 1/2... Discriminator Loss: 1.1898... Generator Loss: 0.8807
Epoch 1/2... Discriminator Loss: 1.3120... Generator Loss: 0.5455
Epoch 1/2... Discriminator Loss: 1.3068... Generator Loss: 0.9092
Epoch 1/2... Discriminator Loss: 1.1532... Generator Loss: 1.5351
Epoch 1/2... Discriminator Loss: 1.1813... Generator Loss: 0.9516
Epoch 1/2... Discriminator Loss: 1.1923... Generator Loss: 0.7579
Epoch 1/2... Discriminator Loss: 1.0159... Generator Loss: 0.9742
Epoch 1/2... Discriminator Loss: 1.0561... Generator Loss: 0.8329
Epoch 1/2... Discriminator Loss: 1.3546... Generator Loss: 0.8323
Epoch 1/2... Discriminator Loss: 1.0935... Generator Loss: 0.7593
Epoch 1/2... Discriminator Loss: 1.1672... Generator Loss: 0.8624
Epoch 1/2... Discriminator Loss: 1.1680... Generator Loss: 1.0068
Epoch 1/2... Discriminator Loss: 1.3668... Generator Loss: 0.8826
Epoch 1/2... Discriminator Loss: 1.0645... Generator Loss: 0.9575
Epoch 1/2... Discriminator Loss: 1.3966... Generator Loss: 0.8794
Epoch 1/2... Discriminator Loss: 1.3373... Generator Loss: 1.2179
Epoch 1/2... Discriminator Loss: 1.2519... Generator Loss: 1.0342
Epoch 1/2... Discriminator Loss: 1.2046... Generator Loss: 1.4718
Epoch 1/2... Discriminator Loss: 1.1696... Generator Loss: 0.9171
Epoch 1/2... Discriminator Loss: 1.1611... Generator Loss: 1.2648
Epoch 1/2... Discriminator Loss: 1.1620... Generator Loss: 0.8458
Epoch 1/2... Discriminator Loss: 1.3698... Generator Loss: 0.4888
Epoch 1/2... Discriminator Loss: 1.3948... Generator Loss: 0.4992
Epoch 1/2... Discriminator Loss: 1.1254... Generator Loss: 0.9150
Epoch 1/2... Discriminator Loss: 1.4058... Generator Loss: 0.7109
Epoch 1/2... Discriminator Loss: 1.2507... Generator Loss: 1.0786
Epoch 1/2... Discriminator Loss: 1.3430... Generator Loss: 0.6637
Epoch 1/2... Discriminator Loss: 0.9287... Generator Loss: 1.2210
Epoch 1/2... Discriminator Loss: 1.0955... Generator Loss: 0.8089
Epoch 1/2... Discriminator Loss: 1.0743... Generator Loss: 0.9530
Epoch 1/2... Discriminator Loss: 0.8346... Generator Loss: 1.3075
Epoch 1/2... Discriminator Loss: 1.4463... Generator Loss: 2.2502
Epoch 1/2... Discriminator Loss: 1.0708... Generator Loss: 1.2390
Epoch 1/2... Discriminator Loss: 1.1027... Generator Loss: 0.7931
Epoch 1/2... Discriminator Loss: 1.2127... Generator Loss: 0.5945
Epoch 1/2... Discriminator Loss: 1.0222... Generator Loss: 0.8182
Epoch 1/2... Discriminator Loss: 1.0974... Generator Loss: 0.7930
Epoch 1/2... Discriminator Loss: 1.2880... Generator Loss: 0.7434
Epoch 1/2... Discriminator Loss: 0.9993... Generator Loss: 1.0322
Epoch 1/2... Discriminator Loss: 1.4865... Generator Loss: 0.3963
Epoch 1/2... Discriminator Loss: 1.4719... Generator Loss: 2.3636
Epoch 1/2... Discriminator Loss: 1.0671... Generator Loss: 0.7542
Epoch 1/2... Discriminator Loss: 1.0475... Generator Loss: 0.8331
Epoch 1/2... Discriminator Loss: 1.0048... Generator Loss: 1.0116
Epoch 1/2... Discriminator Loss: 1.0049... Generator Loss: 0.8778
Epoch 1/2... Discriminator Loss: 1.1342... Generator Loss: 0.9394
Epoch 1/2... Discriminator Loss: 1.0134... Generator Loss: 0.8426
Epoch 1/2... Discriminator Loss: 1.0698... Generator Loss: 1.5572
Epoch 1/2... Discriminator Loss: 1.0725... Generator Loss: 0.6766
Epoch 1/2... Discriminator Loss: 0.8087... Generator Loss: 1.5506
Epoch 1/2... Discriminator Loss: 1.0116... Generator Loss: 0.8264
Epoch 1/2... Discriminator Loss: 1.1730... Generator Loss: 0.6089
Epoch 1/2... Discriminator Loss: 1.2866... Generator Loss: 0.5604
Epoch 1/2... Discriminator Loss: 1.1718... Generator Loss: 0.8073
Epoch 1/2... Discriminator Loss: 1.3942... Generator Loss: 0.5579
Epoch 1/2... Discriminator Loss: 1.1896... Generator Loss: 0.8227
Epoch 1/2... Discriminator Loss: 1.6377... Generator Loss: 1.7554
Epoch 1/2... Discriminator Loss: 1.1972... Generator Loss: 0.6944
Epoch 1/2... Discriminator Loss: 0.6946... Generator Loss: 1.5995
Epoch 1/2... Discriminator Loss: 1.3815... Generator Loss: 0.4941
Epoch 1/2... Discriminator Loss: 1.0636... Generator Loss: 0.8060
Epoch 1/2... Discriminator Loss: 0.9354... Generator Loss: 1.0793
Epoch 1/2... Discriminator Loss: 1.0450... Generator Loss: 0.7753
Epoch 1/2... Discriminator Loss: 1.0569... Generator Loss: 0.7878
Epoch 1/2... Discriminator Loss: 0.8921... Generator Loss: 2.2160
Epoch 1/2... Discriminator Loss: 0.8378... Generator Loss: 1.0973
Epoch 1/2... Discriminator Loss: 1.8029... Generator Loss: 0.2531
Epoch 1/2... Discriminator Loss: 1.2267... Generator Loss: 1.9925
Epoch 1/2... Discriminator Loss: 1.7081... Generator Loss: 0.2915
Epoch 1/2... Discriminator Loss: 0.6059... Generator Loss: 2.0881
Epoch 1/2... Discriminator Loss: 0.8065... Generator Loss: 1.5672
Epoch 1/2... Discriminator Loss: 0.7229... Generator Loss: 1.3082
Epoch 1/2... Discriminator Loss: 1.4483... Generator Loss: 0.4318
Epoch 1/2... Discriminator Loss: 1.2481... Generator Loss: 0.4569
Epoch 1/2... Discriminator Loss: 1.4266... Generator Loss: 0.3909
Epoch 1/2... Discriminator Loss: 1.8027... Generator Loss: 0.2296
Epoch 1/2... Discriminator Loss: 1.5143... Generator Loss: 0.3515
Epoch 1/2... Discriminator Loss: 0.4969... Generator Loss: 1.8650
Epoch 1/2... Discriminator Loss: 0.8662... Generator Loss: 0.8423
Epoch 1/2... Discriminator Loss: 0.9858... Generator Loss: 0.9931
Epoch 1/2... Discriminator Loss: 1.5552... Generator Loss: 1.0613
Epoch 1/2... Discriminator Loss: 0.4099... Generator Loss: 2.0595
Epoch 1/2... Discriminator Loss: 1.0065... Generator Loss: 3.8377
Epoch 1/2... Discriminator Loss: 0.8574... Generator Loss: 2.5452
Epoch 1/2... Discriminator Loss: 0.8916... Generator Loss: 1.6493
Epoch 1/2... Discriminator Loss: 0.6805... Generator Loss: 1.1651
Epoch 1/2... Discriminator Loss: 1.3764... Generator Loss: 0.3762
Epoch 1/2... Discriminator Loss: 0.5394... Generator Loss: 2.2952
Epoch 1/2... Discriminator Loss: 0.6008... Generator Loss: 1.3434
Epoch 2/2... Discriminator Loss: 1.1528... Generator Loss: 0.6930
Epoch 2/2... Discriminator Loss: 0.9571... Generator Loss: 1.0537
Epoch 2/2... Discriminator Loss: 1.7363... Generator Loss: 3.1192
Epoch 2/2... Discriminator Loss: 0.8017... Generator Loss: 2.1613
Epoch 2/2... Discriminator Loss: 1.7865... Generator Loss: 3.2233
Epoch 2/2... Discriminator Loss: 0.6646... Generator Loss: 3.0508
Epoch 2/2... Discriminator Loss: 0.7169... Generator Loss: 1.7816
Epoch 2/2... Discriminator Loss: 0.5380... Generator Loss: 1.8883
Epoch 2/2... Discriminator Loss: 0.5892... Generator Loss: 1.5165
Epoch 2/2... Discriminator Loss: 0.3720... Generator Loss: 2.7054
Epoch 2/2... Discriminator Loss: 0.5002... Generator Loss: 2.0473
Epoch 2/2... Discriminator Loss: 0.4674... Generator Loss: 1.8401
Epoch 2/2... Discriminator Loss: 0.8526... Generator Loss: 0.8924
Epoch 2/2... Discriminator Loss: 0.6970... Generator Loss: 2.7698
Epoch 2/2... Discriminator Loss: 1.2308... Generator Loss: 1.1602
Epoch 2/2... Discriminator Loss: 1.0315... Generator Loss: 0.6796
Epoch 2/2... Discriminator Loss: 1.2815... Generator Loss: 1.7047
Epoch 2/2... Discriminator Loss: 0.6023... Generator Loss: 1.2646
Epoch 2/2... Discriminator Loss: 0.9212... Generator Loss: 0.6662
Epoch 2/2... Discriminator Loss: 1.3002... Generator Loss: 0.5131
Epoch 2/2... Discriminator Loss: 0.7700... Generator Loss: 4.1299
Epoch 2/2... Discriminator Loss: 0.2371... Generator Loss: 2.3717
Epoch 2/2... Discriminator Loss: 0.9650... Generator Loss: 0.7026
Epoch 2/2... Discriminator Loss: 0.8487... Generator Loss: 2.3091
Epoch 2/2... Discriminator Loss: 0.5795... Generator Loss: 1.1231
Epoch 2/2... Discriminator Loss: 2.7579... Generator Loss: 2.3830
Epoch 2/2... Discriminator Loss: 1.2544... Generator Loss: 3.0773
Epoch 2/2... Discriminator Loss: 1.0882... Generator Loss: 0.6692
Epoch 2/2... Discriminator Loss: 0.6360... Generator Loss: 1.1742
Epoch 2/2... Discriminator Loss: 0.4689... Generator Loss: 2.5642
Epoch 2/2... Discriminator Loss: 1.0475... Generator Loss: 2.7829
Epoch 2/2... Discriminator Loss: 0.8000... Generator Loss: 1.7077
Epoch 2/2... Discriminator Loss: 0.3897... Generator Loss: 2.1791
Epoch 2/2... Discriminator Loss: 0.6186... Generator Loss: 4.6576
Epoch 2/2... Discriminator Loss: 0.2431... Generator Loss: 3.9742
Epoch 2/2... Discriminator Loss: 0.2399... Generator Loss: 3.5197
Epoch 2/2... Discriminator Loss: 1.1788... Generator Loss: 0.4499
Epoch 2/2... Discriminator Loss: 0.3986... Generator Loss: 4.4891
Epoch 2/2... Discriminator Loss: 1.1206... Generator Loss: 0.5992
Epoch 2/2... Discriminator Loss: 0.4725... Generator Loss: 1.4065
Epoch 2/2... Discriminator Loss: 1.0093... Generator Loss: 2.7885
Epoch 2/2... Discriminator Loss: 0.3991... Generator Loss: 1.6670
Epoch 2/2... Discriminator Loss: 0.4457... Generator Loss: 1.3410
Epoch 2/2... Discriminator Loss: 2.1672... Generator Loss: 3.2074
Epoch 2/2... Discriminator Loss: 0.3946... Generator Loss: 2.2331
Epoch 2/2... Discriminator Loss: 0.2478... Generator Loss: 2.0838
Epoch 2/2... Discriminator Loss: 0.9971... Generator Loss: 0.5844
Epoch 2/2... Discriminator Loss: 0.2558... Generator Loss: 2.8129
Epoch 2/2... Discriminator Loss: 0.4403... Generator Loss: 1.7383
Epoch 2/2... Discriminator Loss: 0.1678... Generator Loss: 2.7029
Epoch 2/2... Discriminator Loss: 3.1956... Generator Loss: 1.8448
Epoch 2/2... Discriminator Loss: 0.8415... Generator Loss: 1.4017
Epoch 2/2... Discriminator Loss: 0.4664... Generator Loss: 1.4022
Epoch 2/2... Discriminator Loss: 0.5138... Generator Loss: 2.6944
Epoch 2/2... Discriminator Loss: 1.5133... Generator Loss: 2.4772
Epoch 2/2... Discriminator Loss: 0.6063... Generator Loss: 1.5276
Epoch 2/2... Discriminator Loss: 0.3111... Generator Loss: 2.0907
Epoch 2/2... Discriminator Loss: 0.3731... Generator Loss: 1.5645
Epoch 2/2... Discriminator Loss: 0.9539... Generator Loss: 1.4935
Epoch 2/2... Discriminator Loss: 0.5149... Generator Loss: 1.1852
Epoch 2/2... Discriminator Loss: 0.2047... Generator Loss: 3.2649
Epoch 2/2... Discriminator Loss: 0.4443... Generator Loss: 1.4776
Epoch 2/2... Discriminator Loss: 0.7008... Generator Loss: 0.8497
Epoch 2/2... Discriminator Loss: 0.4274... Generator Loss: 3.9244
Epoch 2/2... Discriminator Loss: 0.6661... Generator Loss: 0.9511
Epoch 2/2... Discriminator Loss: 0.2370... Generator Loss: 2.6572
Epoch 2/2... Discriminator Loss: 0.7426... Generator Loss: 0.7759
Epoch 2/2... Discriminator Loss: 1.0626... Generator Loss: 0.6126
Epoch 2/2... Discriminator Loss: 0.3930... Generator Loss: 2.2557
---------------------------------------------------------------------------
KeyboardInterrupt                         Traceback (most recent call last)
/usr/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py in get_controller(self, default)
   3624       self.stack.append(default)
-> 3625       yield default
   3626     finally:

<ipython-input-15-8ad43362037f> in <module>()
     15     train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
---> 16           celeba_dataset.shape, celeba_dataset.image_mode)

<ipython-input-11-779a49d31bd0> in train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode)
     49                 _ = sess.run(net.d_opt, feed_dict={net.input_real: batch_images, net.input_z: batch_z})
---> 50                 _ = sess.run(net.g_opt, feed_dict={net.input_z: batch_z, net.input_real: batch_images})
     51 

/usr/lib/python3.4/dist-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
    777       result = self._run(None, fetches, feed_dict, options_ptr,
--> 778                          run_metadata_ptr)
    779       if run_metadata:

/usr/lib/python3.4/dist-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
    981       results = self._do_run(handle, final_targets, final_fetches,
--> 982                              feed_dict_string, options, run_metadata)
    983     else:

/usr/lib/python3.4/dist-packages/tensorflow/python/client/session.py in _do_run(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)
   1031       return self._do_call(_run_fn, self._session, feed_dict, fetch_list,
-> 1032                            target_list, options, run_metadata)
   1033     else:

/usr/lib/python3.4/dist-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
   1038     try:
-> 1039       return fn(*args)
   1040     except errors.OpError as e:

/usr/lib/python3.4/dist-packages/tensorflow/python/client/session.py in _run_fn(session, feed_dict, fetch_list, target_list, options, run_metadata)
   1020                                  feed_dict, fetch_list, target_list,
-> 1021                                  status, run_metadata)
   1022 

KeyboardInterrupt: 

During handling of the above exception, another exception occurred:

IndexError                                Traceback (most recent call last)
<ipython-input-15-8ad43362037f> in <module>()
     14 with tf.Graph().as_default():
     15     train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
---> 16           celeba_dataset.shape, celeba_dataset.image_mode)

/usr/lib64/python3.4/contextlib.py in __exit__(self, type, value, traceback)
     75                 value = type()
     76             try:
---> 77                 self.gen.throw(type, value, traceback)
     78                 raise RuntimeError("generator didn't stop after throw()")
     79             except StopIteration as exc:

/usr/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py in get_controller(self, default)
   3626     finally:
   3627       if self._enforce_nesting:
-> 3628         if self.stack[-1] is not default:
   3629           raise AssertionError(
   3630               "Nesting violated for default stack of %s objects"

IndexError: list index out of range

Submitting This Project

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.